Engineering & Construction

The question

Are specialty engineering and construction firms priced for the run-rate of backlog already sitting on their balance sheets, or for a mean-reverting cycle that the disclosed book-to-bill ratios contradict?

Base rates

The reference class is prior multi-year non-residential construction cycles: the 1993-1999 telecom and office build, the 2004-2008 commercial and energy build, and the 2011-2015 shale-related civil cycle. In each cohort, specialty mechanical, electrical, and civil contractors compounded revenue 12-18% annually through mid-cycle with operating-margin expansion of 150-300 basis points, then gave back one to two turns of multiple as backlog rolled. The historical base rate for a diversified specialty-contractor basket beating the S&P over a three-year mid-cycle window is roughly the 60th-70th percentile of industrial subsectors. Importantly, the base rate for margin expansion -- not revenue -- is where most of the excess return has historically come from, because labor is the binding constraint and price discipline holds when the backlog is 18-30 months deep.

Consensus forward numbers for this cohort impute that 2027-2028 revenue grows mid-single digits with flat margins. That is a reasonable prior for a normal cycle. It is likely too conservative given that disclosed backlogs are currently extending rather than compressing, and that the mix is shifting toward higher-margin mechanical and data-center scopes.

Why the consensus view is wrong (or incomplete)

The sell-side models this group as a cyclical beta play on non-residential construction spending. That framework treats backlog as a coincident indicator of the cycle. The causal mechanism it misses is that backlog in specialty trades is increasingly pre-committed under cost-plus or guaranteed-maximum-price structures tied to data-center, semiconductor-fab, LNG, and transmission projects with in-service dates in 2027-2029. These are not speculative bids; they are signed contracts with anchor tenants who have already taken the capex on their own balance sheets.

The second missed mechanism is labor. Skilled mechanical labor is the binding constraint on delivery, which means the contractors with training pipelines and regional density are capturing price, not volume. That is a margin story, not a revenue story, and sell-side models underweight it because their historical elasticities were calibrated on a labor-abundant cycle.

The market is pricing cyclical risk. The real risk is execution -- specifically, whether the labor pipeline delivers on the backlog at the margins already booked. That asymmetry is the inefficiency.

Position construction

The book organizes into three sub-books.

The first is mechanical and data-center specialty, anchored by FIX at 20% -- the cleanest pure-play on mechanical contracting with a data-center mix that is expanding faster than disclosed backlog alone suggests. AGX (~4%) captures EPC for gas-fired generation and transmission, a smaller position because its backlog is lumpier and customer-concentrated. IBP (~7.5%) is the residential and light-commercial insulation integrator and plays a counter-cyclical role relative to the heavier industrial scopes.

The second is civil and infrastructure, anchored by MTZ (~17%) for electrical-transmission and pipeline scope, DY (~9%) for telecom-fiber build, PRIM (~8%) for utility and renewables civil, and FLR (~8%) for large-scale EPC optionality where the re-rating is earliest. ROAD (~7%) and GVA (~5%) are state and municipal paving exposure -- lower beta, steadier book-to-bill, funded by multi-year surface-transportation authorizations. WMS (~12%) is the water-management materials franchise; NWPX (~0.5%) is a small position in large-diameter steel pipe where we want exposure but do not want concentration.

The third sub-book is a single small position, DRH (~1.7%), which sits outside the core thesis and reflects a secondary leisure-demand read-through on domestic industrial employment; we flag it as a non-core holding and size it accordingly.

The top three positions concentrate the book in the names where backlog visibility is highest and margin expansion is most likely to persist.

Asymmetric payoff

If backlog converts at the rates disclosed in current investor decks and margins expand 75-150 basis points through 2027, the weighted book returns roughly 20-30% per year. If non-residential construction spending flat-lines and multiples compress one turn, the book returns roughly -10 to -20%. If labor-cost pass-through holds and the data-center mechanical mix expands faster than consensus, the right tail is 35-50% with a re-rating toward quality-compounder multiples.

Assigning a 55% probability to the base case, a 25% probability to the bear, and a 20% probability to the bull produces an expected value in the range of +14 to +20% annualized against an SPY base rate near +8%. The payoff is asymmetric because book-to-bill in mechanical and civil is already booked; the bear case is a multiple compression, not an earnings collapse.

Three things that would change our mind

1. Two consecutive quarters of declining book-to-bill below 1.0x across three or more of the top six holdings, indicating that backlog is rolling off faster than new awards.
2. Operating-margin compression of more than 100 basis points at FIX or MTZ on the quarterly call, which would signal that labor-cost pass-through has broken.
3. A sustained slowdown in data-center and fab announcements below the 2024-2025 run rate for two consecutive quarters, which would weaken the forward bid for specialty mechanical scope.

What we are explicitly NOT betting on

We are not buying residential homebuilders, pure commercial-real-estate exposure, or the heavy-machinery OEMs that sell into this cycle. The unit economics of an OEM are dominated by a shorter order book and more cyclical price-cost dynamics than a specialty contractor's multi-year backlog provides. We are also not betting on a specific federal-policy outcome; the backlog thesis holds across plausible policy configurations because the anchor customers have already taken the capex. The margin of safety comes from signed work, not from a forecast of fiscal generosity.